Table D.1.

Monte Carlo estimates of chance prediction metrics. P-value is the probability of chance explaining the S_n=80% of a matched reference algorithm. The expected values according to the Poisson formulas are E[S_nc] = 0.27241, E[ρ_w] = 0.275, E[r_w] = 0.155/hr, and p = 0.02153.

Seizures vs. Chance Predictor Distributions	Quantity of Interest	Monte Carlo Simulation
Uniform, Poisson	E[Snc]	0.27053
	E[ρw]	0.275
	E[rw]	0.156/hr
	p-value	0.02008
Gamma, Poisson	E[Snc]	0.27390
	E[ρw]	0.275
	E[rw]	0.155/hr
	p-value	0.02480
Gamma, Gamma	E[Snc]	0.27558
	E[ρw]	0.275
	E[rw]	0.131/hr
	p-value	0.03100